Cytosponge Adequacy Study Evaluation II: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cytosponge Adequacy Study Evaluation II (NCT02395471), a non-randomized na study. The primary outcome is Number of Participants With Adequate Cytosponge™ Sample.

2. Clinical Question

The clinical question is about sample collection adequacy, not a comparison of clinical outcomes between treatments. The endpoint description says the study assessed the adequacy of cytology samples obtained after 1 sampling and, when the first sample was inadequate, after 2 samplings. The recorded timeframe is immediately post procedure up to 7 days +/- 3 days.

The two participant descriptions identify different routes into the study: routine endoscopic Barrett’s surveillance for one group, and upper endoscopy for screening for Barrett’s among people with GERD symptoms for the other. These descriptions establish the intended clinical contexts, but do not define a randomized contrast. The record does not report the exact adequacy criteria, the number meeting those criteria, or a comparison with another collection method. Those details are not reported in the registry record.

3. Trial Design

The registry identifies this as a completed, non-randomized, single-group study with no masking. It records an enrollment of 191 and lists the Cytosponge™ Cell Collection Device as the intervention. The study is titled Cytosponge Adequacy Study Evaluation II, and its official title describes assessment of a minimally invasive esophageal cytology collection system in patients with Barrett’s esophagus or GERD symptoms.

The record describes two clinical populations: people with Barrett’s esophagus presenting for routine endoscopic surveillance and people with GERD symptoms undergoing upper endoscopy for screening for Barrett’s esophagus. These are not randomized treatment arms. The record’s design classification is single-group, and it does not specify a separate control intervention. The populations may differ in why they underwent endoscopy, so a difference between their results, if one were reported, would not by itself establish a causal group effect.

The study start is recorded as 2015-08, and primary completion as 2018-04-24. The record lists the status as completed and the phase as NA. It names Medtronic - MITG as sponsor. These administrative details do not supply information about how participants were selected, whether all enrolled participants completed collection, or how the endpoint was analyzed.

Non-randomized and unmasked designs can be appropriate for describing whether a procedure yields an adequate sample, but they do not provide the protections of randomized allocation or masking against all sources of bias. Here, the design is primarily descriptive as recorded: the key question is whether collected samples were adequate, rather than whether the device outperformed a control. The registry’s primary outcome result, eligibility details, and analysis plan are not reported in the registry record. The listed enrollment should therefore not be mistaken for the number with an observed primary outcome.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Adequate Cytosponge™ Sample, measured over Immediately post procedure up to 7 days +/- 3 days.

The primary endpoint is the number of participants with an adequate Cytosponge™ sample. It is a participant count: each participant is classified according to whether the sample meets the study’s adequacy criterion. The endpoint description specifies assessment after 1 sampling and after 2 samplings if the first sample is inadequate. The recorded assessment window is immediately post procedure up to 7 days +/- 3 days.

Sample adequacy is a practical measure of whether the collection procedure produced material suitable for the intended cytology assessment. For a collection-system evaluation, this helps distinguish successful acquisition of an evaluable sample from an unsuccessful or insufficient collection. It is not, by itself, a measure of diagnostic accuracy, disease severity, or clinical benefit. Those questions require their own definitions and results.

The registry names the endpoint but does not report the number of participants whose samples were adequate, the denominator used for the endpoint, or the criterion used to decide adequacy. It also does not report how participants who needed a second sampling were counted in the final endpoint. These details are not reported in the registry record. Without the numerator, denominator, and operational definition, the posted record does not support an estimate of the observed adequacy proportion or its precision.

In general, a participant count is easier to interpret alongside the number assessed. The count alone does not indicate how common an outcome was if the relevant denominator is unknown. Likewise, a confidence interval around a proportion would communicate the uncertainty associated with the observed data, but no primary adequacy estimate or interval is reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry labels the primary outcome as a count of participants. For a binary outcome, each participant is classified into one of two states, such as meeting or not meeting a specified adequacy criterion. The count of participants meeting the criterion is the numerator of a proportion when divided by the relevant number assessed. That proportion is often easier to compare across studies than a raw count, but the denominator must be clearly defined. The primary endpoint result and denominator are not reported in the registry record, so neither a proportion nor its uncertainty can be described.

What does a confidence interval add?

In general, a confidence interval expresses the statistical uncertainty around an estimate. For a binary endpoint, an interval around the proportion would help show how compatible the observed data are with a range of underlying adequacy frequencies. A narrow interval generally indicates greater precision than a wide interval, although precision is not the same as freedom from bias. An interval cannot correct for selection differences, outcome misclassification, or a missing comparison group. No primary endpoint interval is reported in the registry record.

What can the design establish?

The study is listed as non-randomized, unmasked, and single-group. Random allocation is a general method for balancing measured and unmeasured participant characteristics between treatment groups, on average. Since randomization is not recorded and no separate control intervention is specified, the results cannot be interpreted as a randomized comparison of the device against an alternative. The Barrett’s and GERD descriptions identify different clinical populations, not treatment assignments. Descriptive outcome information could characterize sample adequacy in the enrolled population, but a difference between these populations would not isolate the effect of a device or establish why the groups differed.

Masking is also listed as none. In general, masking can reduce the influence of knowledge about a procedure or assignment on assessment and reporting. Whether masking would be feasible for this collection procedure, and whether any independent adequacy assessment was used, is not reported in the registry record. The absence of recorded masking is a design feature to consider, not evidence that an assessment was biased.

What do the planning assumptions mean?

The registry’s planning assumptions describe an illustration based on the observed effect, not the trial’s own design. They specify a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. These assumptions are not evidence that participants were randomized or allocated equally in the study. They are planning conventions for an illustrative sample-size calculation, distinct from the actual single-group, non-randomized design.

In general, a significance level is a threshold used in a specified statistical test to control the probability of a false positive under the test’s assumptions. A two-sided test allows evidence in either direction to count against the null hypothesis. A normal approximation uses a mathematical distribution to approximate the sampling behavior of an estimate or test statistic. Such approximations can be convenient, but their adequacy depends on the data and the calculation being used. The registry does not report the trial’s own statistical analysis plan, so it is not possible to say whether these assumptions were used for the trial’s outcome analyses.

Planning for a fraction of an observed effect is a way to consider what sample size might be needed if a future or illustrative study aimed to detect a smaller effect than the one used as a reference. It does not mean that the trial had that target, nor does a calculated sample size show that a study had a particular probability of finding an effect. Sample-size planning and observed evidence answer different questions: planning concerns the design assumptions, while an estimate and its uncertainty concern the data actually collected. The planning assumptions do not supply the missing primary endpoint result.

Why does over-dispersion matter in general?

For count outcomes, some statistical models assume that variability follows a particular relationship to the average count. Over-dispersion means the observed variation is greater than that model expects. If ignored, standard errors may be too small and intervals or tests may appear more decisive than warranted. This is a general issue for count data, not a reported finding in this study. The registry classifies the primary endpoint as a participant count and does not report a model, a dispersion estimate, or an analysis addressing over-dispersion.

The endpoint is described as a count of participants with an event, rather than a continuous measurement. General methods for continuous outcomes, such as comparing means, would not automatically be appropriate for a binary participant-level endpoint. A valid analysis would need to state the endpoint definition, analysis population, denominator, handling of repeated sampling, and method for quantifying uncertainty. Those details are not reported in the registry record. The record also lists 0 analyses, which means no analysis entries are recorded there; it does not establish that no analyses were conducted outside the registry.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Patient Wth Barrett's", "Patients With GERD")
n      <- c(127, 61)
events <- c(115, 56)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Patient Wth Barrett's  127  115  90.6  84.1  95.0  84.2  94.5  84.2  94.5
ARM  Patients With GERD  61  56  91.8  81.9  97.3  82.2  96.4  82.2  96.4

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 7 outcome measures and names secondary topics covering operating characteristics, comparisons with worst histology ever, operating characteristics by baseline histology, endoscopic observations of abrasion, bleeding, and perforation, and ongoing safety measures. It records 0 analyses. These entries do not give a complete account of how many statistical tests were conducted, which outcomes were prioritized in the analysis, or how results were selected for reporting.

When several outcomes or subgroup analyses are examined, the chance of at least one apparently unusual result can increase. In general, multiplicity procedures can adjust how statistical evidence is assessed across a set of tests, or investigators can specify a hierarchy of outcomes in advance. Whether any such approach was planned or used is not reported in the registry record. The listed outcome topics alone do not show that multiplicity adjustment was necessary, performed, or omitted in an analysis.

The registry does not report how missing or inadequate samples were handled in the primary endpoint calculation, whether all enrolled participants were included in each analysis, or how exclusions and unavailable assessments were addressed. The exact analysis populations and missing-data methods are not reported in the registry record. These details matter because a count based only on participants with completed assessments can differ in interpretation from a count based on all enrolled participants.

The registry’s analysis count of 0 should be read as absence of recorded analysis entries, not as proof that no analysis was performed. Likewise, the absence of a posted primary adequacy result does not establish that the endpoint was unmeasured. The registry record does not provide enough information to determine the analysis population, missingness, or multiplicity strategy.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: sponge detachment (1.6%).

For the Barrett’s group, the registry records 2 participants with serious adverse events among 129 at risk and 8 with other adverse events among 129 at risk. For the GERD group, it records 0 with serious adverse events among 62 at risk and 6 with other adverse events among 62 at risk. Deaths are listed as 0 for each group. The registry also names Sponge Detachment at 1.6% among the top serious terms.

These are registry counts and denominators, not a randomized comparison. The two groups differ by the clinical context described in the record, and no like-for-like control group is specified. The counts should not be interpreted as evidence that the device caused or prevented an event. They also do not, by themselves, indicate how events were defined, how long participants were monitored, or whether all events were captured; those details are not reported in the registry record.

Serious adverse events and other adverse events are separate categories in the record. The term Sponge Detachment is listed among top serious terms, but the record does not provide further detail about its circumstances or attribution. These figures describe the safety information posted, rather than establishing a causal relationship or a complete safety profile.

13. What the Result Does, and Does Not, Mean

The registry record establishes that the study evaluated Cytosponge sample adequacy in people with Barrett’s esophagus or GERD symptoms and records a primary endpoint focused on the number of participants with an adequate sample. It also records the timing of assessment as immediately post procedure up to 7 days +/- 3 days. However, the primary adequacy result is not reported in the registry record. Therefore, the posted information does not establish how often adequate samples were obtained or how precise that estimate would be.

The record’s enrollment of 191 describes enrollment, not the number with a completed and evaluable primary outcome. The endpoint denominator and the adequacy criterion are not reported. It is not possible to infer either from the enrollment figure. The endpoint description includes a second sampling when the first sample is inadequate, but the record does not say how this pathway was incorporated into the final count.

The record also lists operating-characteristic and safety outcome topics, but titles are not outcome results. No numerical primary result or detailed diagnostic-performance result is available in the registry information. The safety counts are descriptive and do not establish causation. Because the study is non-randomized, unmasked, and single-group, and no separate control intervention is specified, it does not provide a randomized treatment comparison.

In general, a study can be informative about whether a collection procedure yields evaluable material without answering every question about diagnostic accuracy, clinical consequences, or comparative performance. Those broader interpretations require appropriately defined outcomes and reported estimates. In this case, the missing adequacy result prevents a quantitative assessment of the study’s primary question. The posted information should be read as a record of the study design, endpoint definition, and some safety counts, not as a complete account of the evidence for all outcomes named in the registry.

14. Why This Trial Matters Statistically

Sample adequacy is a statistically distinct question from whether a diagnostic test correctly classifies disease. A collection system must first yield material that can be assessed; an adequacy endpoint can describe that practical step. Separating collection success from later diagnostic performance helps prevent a count of adequate samples from being interpreted as an accuracy estimate or a patient outcome.

This study is also a useful example of why design labels matter. A registry can describe more than one clinical population while still classifying the study as single-group and non-randomized. Those population descriptions do not create a valid control comparison. Readers need to distinguish participant groups from treatment arms and avoid attributing any group difference to an intervention without a design that supports that inference.

Registry reporting is equally important to interpretation. A primary endpoint title identifies the question, but a result needs a numerator, a denominator, and enough information about outcome definitions and analysis to be interpretable. Without the primary adequacy result, the study’s central quantitative question cannot be answered from the record. The recorded safety counts provide some descriptive information, but do not replace the missing primary outcome.

Finally, planning calculations and observed evidence should be kept separate. Planning assumptions can illustrate what a hypothetical design might require under stated assumptions. They do not describe the actual allocation, prove that a design had adequate power, or provide the missing observed result. This distinction is especially relevant where the stated planning illustration assumes equal allocation between two arms but the study itself is recorded as non-randomized and single-group.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary limitation for interpretation is that the registry names the primary endpoint but does not report its result. The endpoint denominator, observed count, adequacy definition, and handling of second sampling are also not reported in the registry record. As a result, the central adequacy question cannot be quantified from the posted information.

The study is recorded as non-randomized, unmasked, and single-group, with no separate control intervention specified. The Barrett’s and GERD populations have different clinical descriptions and should not be treated as interchangeable comparison groups. Any unadjusted contrast between them would not by itself establish a device effect. The record does not report whether adequacy was assessed independently or whether the assessment was masked.

The registry lists outcome topics but does not give their numerical results in the information available for this analysis. It records 0 analyses, but the statistical analysis plan, analysis populations, multiplicity approach, and methods for handling missing assessments are not reported. The study’s enrollment of 191 does not resolve the absent endpoint denominator.

The safety information is limited to recorded counts by group, deaths, and a listed serious-event term. Further detail on event definitions, attribution, monitoring, and completeness is not reported in the registry record. These limitations restrict what can be concluded from the posted information; they do not show that unreported analyses or assessments were never conducted.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Cytosponge Adequacy Study Evaluation II — NCT02395471.
• Registry record last updated 2019-07-30; status last verified 2019-07.
• Results first posted 2019-07-30.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT02395471 is a completed, non-randomized, unmasked, single-group device study with an enrollment of 191. It evaluated the Cytosponge™ Cell Collection Device in participants with Barrett’s esophagus undergoing routine surveillance or GERD symptoms undergoing upper endoscopy for screening for Barrett’s esophagus. The study start is recorded as 2015-08, and primary completion as 2018-04-24.

The primary endpoint is the number of participants with an adequate Cytosponge™ sample, assessed after 1 sampling and after 2 samplings if the first sample was inadequate. The primary adequacy result and its denominator are not reported in the registry record. Safety counts are recorded separately for the Barrett’s and GERD groups, with 0 deaths in each group. The record lists 7 outcome measures and 0 analyses; detailed statistical and missing-data methods are not reported.